Development of telecommunication system units in parallel neural network basis

Abstract

The article suggests the integration of a neural network as a parallel element base in a telecommunication system. In this case, the ability to learn or adapt to external conditions is applied as the main advantage. For telecommunication systems in conditions when it is possible, this ability will improve noise immunity, reliability, operability, etc. The article considers an example of the integration of a neural network into a discrete matched signal filter. It is noted that the use of parallel mathematical methods in signal processing leads to the maximum effect of increasing the quality parameters of such telecommunication elements